Original Article

CHEMICAL PROPERTIES

Significant work has been focused on understanding the chemical features of sphingolipids based on the belief that the functional biochemistry of these compounds can be better understood. Sphingolipids are, by definition, those compounds that contain a long-chain sphingoid base. The most common long-chain bases are C-18 and C-20 sphingosines⁴. These bases contain a C4-C5 double bond in the trans D-erythro conformation. C-18 and C-20 sphinganines (dihydrosphingosines) are the next most common long-chain bases. These lack the double bond.

When a fatty acid is linked to the amide group of the long-chain base, the resultant lipid is referred to as ceramide. The fatty acids vary in composition, but they are typically long. The most common fatty acids are palmitic (C16:0) and stearic (C18:0). -Hydoxylated fatty acids are often present as well.

The primary alcoholic function at carbon one of ceramide is the site of attachment for a variety of additional moieties. These most commonly include phosphocholine-producing sphingomyelins and saccharides producing a multitude of glycosphingolipids. When a single monosaccharide is present (glucose or galactose), the lipid is referred to as a cerebroside. There may be as many as 15 to 20 sugars in the oligosaccharide chain. When these sugars are uncharged, the products are referred to as neutral glycosphingolipids. Acidic glycosphingolipids result when one or more sialic acids are present (gangliosides) or when a sulfate group is present (sulfatides)⁵. Less commonly, a phosphate may be placed at the C-1 carbon-producing ceramide-1-phosphate⁶. Recently, it has been shown that an additional acyl group may be present at the C-1 position as well producing 1-O-acylceramide⁷.

The ceramide portion of a sphingolipid is highly hydrophobic. When a polar head group is present, such as a phosphorylcholine or oligosaccharide, the sphingolipid becomes amphipathic in character. The presence of the phosphorylcholine head group on both phospatidylcholine and sphingomyelin would lead one to conjecture that physicochemical properties are similar because of a similar molecular conformation. However, there are pronounced differences between the two lipids. First, differences in the number of methylene groups (parafinnic tail length) are more pronounced between the long-chain base and fatty acids in ceramide than in the diglyceride function of phosphatidylcholine. These parafinnic tails are believed to interdigitate with like molecules on either side of the lipid bilayer⁸. In addition, the interfacial region at the juncture of the lipid bilayer and the polar head group of sphingomyelin is believed to be more polar than the comparable region in phosphatidylcholine. This creates more potential sites of interaction between water and other compartments of the lipid bilayer, including the phosphate, free hydroxyl at C-3, and the trans double bond at C4 and C5⁹. The consequences of these chemical interactions include a higher melting point, increased microviscosity, enhanced stability, and decreased permeability to electrolytes than is observed with glycerolipids of homologous structures¹⁰.

The compositional diversity of membrane lipids is thought to facilitate a lateral phase separation of different lipid components. In the case of sphingolipids, the decrement in membrane fluidity may cause microdomains or clusters of lipids rich in ceramide to form. This separation may be facilitated by the composition of the polar head groups on both neutral and acidic glycosphingolipids. Factors that further promote clustering include the content of sialic acid¹¹, the presence of Ca²⁺ ions¹², the ceramide content¹³, and proteins such as VIP21 (caveolin)¹⁴. The theoretical ability of sphingolipids to create distinct microenvironments within the cell membrane has led investigators to focus on these sites as "hot spots" for transmembrane signaling sites, as well as sites of vesciculation and membrane transport¹⁵.

Several acronyms have been applied to these theoretical membrane domains. These include "detergent-resistant membranes," "detergent insoluble glycolipid-enriched membrane domains," and "lipid rafts"¹⁶. Small tubovesicular invaginations within the plasma membrane, termed caveolae, are included within detergent-resistant fractions. In addition, density gradient methods can be used to isolate caveolin-containing membrane fractions that are enriched in glycosphingolipids¹⁷. On the other hand, glycosphingolipid-enriched fractions can be isolated from cells not containing caveolin or having morphologic evidence of caveolae¹⁸.

Some investigators have argued, however, that such domains are unlikely to be present. The interactions required to form lateral domains between lipid species are not sufficiently strong or long lived to be biologically significant¹⁹. In addition, the energy required to demix lipid species once aggregated into domains is small. While microscopy-based experimental evidence exists in support of such lipid rafts, a number of critical questions need to be addressed to provide a more cogent model of their biological significance²⁰. These questions include a determination of the size, composition, and dynamics of these structures, as well as the relationship of such rafts to the cytoskeleton and understanding how the outer membrane leaflet that is enriched in putative raft components (glycosphingolipids, cholesterol, sphingomyelin, and GPI-linked proteins) interacts with the inner leaflet of the membrane.

SPHINGOLIPID PATHWAYS

De novo formation of long-chain sphingoid bases begins with the condensation of palmitoyl-CoA and serine, forming 3-ketodihydrosphingosine Figure 1²¹. This pyridoxal 5'-phosphate–dependent enzyme demonstrates distinct specificities for various acyl-CoAs, with palmitoyl-CoA preferred over stearoyl-CoA by a ratio of 5:1²². This specificity is reflected in the relative abundance of 18 and 20 carbon long-chain bases found in mammalian sphingolipids. The reduction to sphinganine (dihydrosphingosine) is then catalyzed by an oxidoreductase, which is NADPH dependent²³. Dihydroceramide (N-acylsphinganine) is formed by an acylation reaction that uses acyl-CoAs as substrates. The enzyme catalyzing the acylation of sphinganine is acyl-CoA:sphingoid base N-acyltransferase. This enzyme is probably present in many different forms, differing in substrate preference and tissue distribution. The fatty acyl-CoA preference as characterized to date is stearoyl-CoA > lignoceryl-CoA > palmitoyl-CoA > oleyl-CoA²⁴.

Figure 1.

Long-chain base synthesis resulting in ceramide formation.

Full figure and legend (18K)

Ceramide is formed through the de novo synthetic route by the unsaturation at the 4,5 positions of the sphingoid amine by another oxidoreductase²⁵. This reaction is nicotinamide adenine dinucleotide (NAD) dependent. Ceramide may alternatively be synthesized by the acylation of sphingosine by an acyl-CoA:sphingoid base N-acyltransferase. Whether this is the same enzyme that acylates dihydrosphinganine has not been established. There does not appear to exist alternative routes of de novo sphingosine synthesis other than via the acylation of sphinganine, unsaturation of dihydroceramide, and deacylation of ceramide by one of several ceramidases. However, long-chain base acylation may, in theory, occur through two alternative routes. First, an acyl-CoA independent, ATP-dependent mechanism may occur through the reversal of ceramidase²⁶. Second, platelet-activating factor may serve as the substrate for N-acetylation²⁷.

Ceramide is a key intermediate in the formation of most of the biologically important sphingolipids Figure 2. Aside from its deacylation to form free sphingosine, these products use the C1 hydroxyl group on ceramide as the site for additional groups. These products include sphingomyelin, cerebrosides and other glycosphingolipids, ceramide-1-phosphate, ceramide-phosphoethanolamine, and 1-O-acylceramide.

Figure 2.

Alternative pathways for ceramide metabolism.

Full figure and legend (20K)

Sphingomyelin was originally proposed by Sribney and Kennedy to be formed by the transfer of CDP-choline to ceramide²⁸. This pathway has not been subsequently confirmed. Instead, the majority of sphingomyelin is thought to synthesized by the transfer of phosphocholine from phosphatidylcholine to ceramide²⁹. The responsible enzyme is sphingomyelin synthase. This enzyme has been localized to the cis and medial Golgi membranes³⁰. However, a second enzyme with similar activity has been detected in the plasma membrane³¹. Because the sphingomyelin synthase pathway may regulate levels of the precursor ceramide and of a product, diacylglycerol, the regulation of these sphingomyelin synthases and sites of action may have important implications for cell-signaling phenomena, as detailed later in this article.

Two additional pathways for sphingomyelin synthesis may exist. First, a two-step pathway involving the formation of ceramide phosphoethanolamine followed by its methylation has been described³². This appears to be a relatively minor pathway. Second, lyso-sphingomyelin (sphingosylphosphorylcholine) may be acylated directly³³. Sphingosylphosphorylcholine has been demonstrated to have cellular actions akin to sphingosine-1-phosphate³⁴. However, there is limited evidence in support of the existence of this lysolipid in mammalian cells. Recently, a bacterial N-deacylase with activity against sphingomyelin has been reported³⁵. A comparable activity in mammalian tissues has not been identified.

Glucosylceramide is formed from ceramide and UDP-glucose by UDP-glucose:ceramide glycosyltransferase³⁶. Glucosylceramide is the precursor for most of the mammalian glycosphingolipids. The reaction occurs on the cytoplasmic side of the cis Golgi³⁷. Subsequent glycosylation steps occur within the luminal side of the cis Golgi. How ceramide is translocated from the endoplasmic reticulum to the Golgi and how glucosylceramide is flipped into the luminal side of the Golgi is unknown. Glucosylceramide can then be further glycosylated within the Golgi or can be directly transported to the plasma membrane. The latter pathway may occur through the action of a glycolipid transfer protein.

Ceramide may also undergo phosphorylation at the C-1 position, forming ceramide-1-phosphate³⁸. This pathway is catalyzed by a distinct calcium-dependent ceramide kinase. Ceramide-1-phosphate may also be formed by the action of bacterial diglyceride kinase³⁹. This latter reaction serves as the basis for the analytical determination of ceramide levels. To date, ceramide kinase activity has been identified in synaptic vesicles⁶, leukemia cells³⁹, and neutrophils⁴⁰.

Another recently described pathway for ceramide involves its acylation at the C-1 position forming 1-O-acylceramide. The substrate for this acylation is phosphatidylethanolamine or phosphatidylcholine⁷. This reaction is catalyzed by a microsomal phospholipase A2 that transacylates the fatty acid⁴¹. The phospholipase is unique in that it is calcium independent and has an acidic pH optimum.

The main catabolic pathways for sphingolipids in general and for ceramide in particular appear to be through endosomal/lysosomal pathways. Glycosphingolipids are degraded via one or more glycosidases, and ceramide is degraded by an acidic ceramidase. The latter reaction results in the formation of sphingosine and a fatty acid. Free sphingosine is released from the lysosome by an uncharacterized pathway. Free sphingosine is further catabolized outside of the lysosome. It may be reacylated forming ceramide or phosphorylated by sphingosine kinase to form sphingosine phosphate⁴². Sphingosine kinase phosphorylates the C-1 position of either sphingosine or sphinganine. It has been characterized as both a cytosolic enzyme and as membrane bound.

The terminal catabolism of sphingosine involves the action of sphingosine-1-phosphate lyase, which degrades sphingosine-1-phosphate to form ethanolamine phosphate and a fatty aldehdye. C-18 sphingosine is degraded to trans-2-hexadecenal, and C-18 sphinganine is degraded to palmitaldehdye⁴³. Alternatively, sphingosine-1-phosphate may be dephosphorylated by the ceramide-1-phosphate phosphatase⁴⁴.

Another pathway for sphingosine metabolism is N-methylation by sphingosine N-methyltransferase⁴⁵. Both monomethyl and dimethyl sphingosines may be formed by this pathway.

REGULATION OF METABOLISM

The subcellular sites of metabolism and the subcellular distribution of sphingolipids are important determinants in understanding the cellular biology of these lipids. The role of various sphingolipids in cell-signaling events has been difficult at times to reconcile with some of the known properties and with the subcellular localization of the relevant enzymes. In addition, it is unknown whether many of these enzymes that colocalize to specific cellular organelles are present in multienzyme complexes.

In general, all of the enzymes associated with de novo long-chain base synthesis and ceramide formation are present in the smooth endoplasmic reticulum⁴⁶. Furthermore, the active sites for these enzymes appear to be on the cytosolic face of the endoplasmic reticulum. The localization and site of the dihydroceramide oxidoreductase have not been established. Acyl-CoA:sphingoid base acyltransferase has also been shown to localize to mitochondria⁴⁷. Whether this mitochondrial activity may have some role in ceramide-induced apoptosis (discussed later in this article) is an intriguing yet unanswered question.

Sphingomyelin biosynthesis appears to occur primarily within the cis and medial Golgi membranes, with the active site within the luminal leaflet. This implies that de novo synthesized ceramide must be transported from the endoplamic reticulum to the Golgi. In addition, there must be a translocation of ceramide across the lipid bilayer. In addition, a similar enzymatic activity has been identified in the plasma membrane⁴⁸. This latter finding implies that ceramide formed from sphingomyelin hydrolysis may be resynthesized at the site of its degradation within the plasma membrane without intracellular transport to the Golgi.

The degradation of sphingomyelin is catalyzed by a phosphodiesterase-producing phosphorylcholine and ceramide. This sphingomyelinase activity is separable into at least seven distinct activities. These include (1) an acidic, lysosomal sphingomyelinase⁴⁹, (2) a zinc-dependent serum sphingomyelinase⁵⁰, (3) a neutral, magnesium-dependent, membrane associate sphingomyelinase⁵¹, (4) a neutral Mg²⁺-independent sphingomyelinase identified in the myelin sheath⁵², (5) a magnesium and dithiothreitol-stimulated neutral sphingomyelinase found in the nuclei of ascites cells⁵³, (6) a neutral sphingomyelinase localized to the chromatin and nuclear envelop of liver cells⁵⁴, and (7) an alkaline sphingomyelinase in the bile and intestine⁵⁵.

Many of these sphingomyelinase activities have been implicated in cellular sphingolipid-signaling events. Thus, considerable attention has been focused on the biochemistry of these lipases in recent years. Structurally, the best studied of these sphingomyelinases is the acid, lysosomal sphingomyelinase⁵⁶. This lipase is deficient in patients with Niemann-Pick disease, a lysosomal storage disorder. The zinc-dependent serum sphingomyelinase has been determined to be a product of the same gene⁵⁷. A neutral, magnesium-dependent, glutathione-stimulated enzyme has recently been sequenced⁵⁸. Whether this lipase is a signaling sphingomyelinase is uncertain.

Both sphingomyelin and glycosphingolipids are highly enriched in the outer leaflet of the plasma membrane. The mechanism whereby sphingolipids flip from the outer to the inner leaflet is unknown. One hypothesis is that the multidrug resistance protein serves as a membrane "flipase" for placement of glycosphingolipids such as glucosylceramide in the outer leaflet⁵⁹. It is noteworthy that tumor cells that express high levels of multidrug resistance protein are characterized by high contents of glucosylceramide⁶⁰.

A SIGNALING ROLE FOR SPHINGOLIPIDS

Interest in the role of sphingolipids as signaling molecules began after the discovery that sphingosine could act as an inhibitor of protein kinase C⁶¹. In 1989, Okazaki, Bell, and Hannun reported the first example of agonist-coupled degradation of sphingomyelin⁶². Since that time, the number of published studies on sphingolipid-based signaling has increased exponentially. While the majority of attention has been focused on ceramide, the product of sphingomyelin hydrolysis, as the principle sphingolipid mediator of intracellular signaling events, several sphingolipid metabolites generated both as a result of sphingomyelinase activity and through de novo synthesis have been implicated in signaling events. Sphingolipids implicated in signaling events include ceramide⁶³, sphingosine⁶⁴, sphingosine-1-phosphate⁶⁵, and ceramide-1-phosphate⁶⁶. The primary proposed signaling functions for these sphingolipids are outlined in Figure 3.

Figure 3.

Some proposed signaling functions of ceramide and its metabolites.

Full figure and legend (16K)

The general picture that has emerged of sphingolipid signaling is both exciting in its implications and confusing in both its complexity and, at times, inconsistencies. In some cases, the implication that sphingolipids such as ceramide may play a primary role in cell signaling has been highly controversial. This has been particularly true for hypotheses stating that ceramide regulates nuclear factor-B (NF-B) signaling and apoptosis^67,68.

In general, those agonists that increase cell ceramide levels cause cell growth arrest, differentiation, or apoptosis. These include the cytokines⁶⁹, environmental stressors⁷⁰, and pharmacologic agents⁷¹. The downstream cellular responses are most often consistent with these effects and include NF-B activation⁷², cyclooxygenase transcription⁷³, heat-shock protein transcription⁷⁴, selectin expression⁷⁵, and cytokine secretion⁷⁶. Conversely, the growth factors are often associated with an increase in sphingosine and sphingosine-1-phosphate⁷⁷. These opposing activities have led Cuvillier to propose a "yin yang" model of intracellular sphingolipids with ceramide and sphingosine-1-phosphate playing opposing roles⁷⁸. In this model, a sphingolipid "rheostat" may exist within cells, the setting or balance of which determines whether cells undergo a proliferative or differentiation of apoptotic response.

Four primary observations in support of a role for sphingomyelinase-stimulated ceramide formation as a transduction event: First, several extracellular stimuli trigger ceramide generation and sphingomyelin degradation. Multiple activators have been associated with the generation of cellular sphingolipids. An incomplete listing would include cytokines [tumor necrosis factor- (TNF-)⁷⁹, interleukin (IL)-1⁸⁰, IL-2⁸¹, and -interferon⁸²], nerve growth factor⁸³, hormones (glucocorticoids⁸⁴, progesterone⁸⁵, 1,25 (OH)₂ vitamin D₃⁸⁶, bradykinin⁸⁷, retinoic acid⁸⁸, endothelin 1⁸⁹), antigens (FAS ligand⁹⁰, anti-CD28⁹¹), environmental stressors (ionizing radiation⁹², heat shock⁷⁴, ischemia⁹³, hypoxia⁹⁴, peroxide⁷⁰), pharmacologicals (daunorubicin⁹⁵, vincristine⁹⁶, Ara-C⁹⁷, phorbol esters⁹⁸, calcium channel blockers⁹⁹, protein kinase C inhibitors¹⁰⁰), and infective agents (viruses¹⁰¹, bacteria¹⁰², and bacterial toxins¹⁰³).

Second, sphingomyelinase activity increases in a kinetically consistent manner. The degree of increase in ceramide is variably reported, ranging from a 20% increase to a 2000% increase^90,94. Concurrent measurements of sphingomyelin and ceramide have been made in many studies. In general, sphingomyelin levels or radiolabeling decrease over time and at a magnitude that is consistent with the appearance of ceramide. Often, the sphingomyelin levels return to control values as the ceramide content decreases, implying that the ceramide is recycled to sphingomyelin, presumably via the sphingomyelin synthase. There is some evidence in support of a membrane-associated sphingomyelin synthase on the inner leaflet of the plasma membrane. There are three implications for this model. First, sphingomyelin hydrolysis, ceramide formation, and sphingomyelin resynthesis would necessarily occur at this site. Second, the ceramide generated would act at this site on a local target or be transported from this site to one where such targets exist. The ceramide would then be reutilized, or ceramide synthesized de novo would serve as substrate for resynthesis. Third, a significant amount of diacylglycerol, the product of sphingomyelin synthesis, would be generated and thus may have independent biological activity.

Not only is the degree of ceramide elevation highly variable, but the time course for ceramide accumulation is as well. Four patterns of ceramide accumulation have been described. These include a rapid increase occurring within a few minutes and declining to baseline within 15 to 20 minutes¹⁰⁴, a gradual increase occurring within an hour and decreasing within two to three hours¹⁰⁵, a rapid increase with a persistent elevation lasting several hours⁷⁶, and a gradual increase occurring after several hours and persisting for 24 hours or longer¹⁰⁶. Surprisingly, these patterns can be observed with the same agonist (IL-1)^{76,104,105,106,107} and are not associated with the degree of rise in ceramide.

A third observation in support of a role for sphingomyelin hydrolysis is that the biological consequences of sphingomyelin degradation and ceramide generation can often be replicated by maneuvers that augment or inhibit ceramide levels. These methods include the addition of short chain, cell-permeant ceramides¹⁰⁸, the addition of bacterial sphingomyelinase¹⁰⁹, the use of inhibitors of long-chain base acylation¹¹⁰, the use of inhibitors that decrease ceramide utilization through glycosylation pathways¹¹¹, the use of inhibitors of ceramide degradation¹¹², or the addition of natural long-chain ceramides¹¹³. Each of these methods is associated with potential problems confounding the interpretation of the experimental data.

Short-chain ceramides, usually containing an acetyl group or C6 or C8 fatty acid at the amide linkage, have been extremely popular in signaling studies. These analogues retain the D-erythro stereospecificity of natural ceramides as well as the critical C4-C5 double bond. In addition, in early studies, these compounds were believed to be relatively inert with regard to cellular metabolism. Thus, any biological activities observed with these compounds were interpreted as mimics for the activities of endogenous ceramides.

Several difficulties exist with this interpretation. First, these ceramides are quite avidly metabolized by cells¹¹⁴. Original studies often used analogues radiolabeled within the fatty acid. The liberation of this fatty acid by a ceramidase would result in a long-chain base, which was unlabeled and in which the metabolism was therefore impossible to follow. More recent studies have demonstrated that following the addition of these compounds, both short- and long-chain sphingomyelins and glycosphingolipids are synthesized as well as free sphingosine, long-chain ceramides, and long-chain sphingomyelins. Another problem is that the concentrations of ceramides added to cells (typically in the micromolar range) result in intracellular levels of ceramides that are approximately 100 times those found normally. Additionally, in some cases, comparable biological effects are observed, irrespective of the stereoselectivity of the analogue employed. Finally, because of distinct physiochemical differences in these analogues, short-chain ceramides may destabilize membranes through detergent-like effects that are independent of their endogenous biological activities¹¹⁵.

The use of exogenous sphingomyelinase has also been problematic, because the majority of sphingomyelin is in the outer leaflet of the plasma membrane where it would be accessible to the bacterial enzyme. However, the pool of sphingomyelin that is susceptible to agonist-stimulated sphingomyelinase activity is believed to exist on the inner membrane leaflet¹¹⁶. Thus, ceramide produced with the bacterial enzyme would be present at a site that may have little to do with the endogenous signaling mechanisms. It is therefore not surprising that the use of this approach often does not replicate the biology under investigation.

Inhibitors of glucosylceramide formation (1-phenyl-2-decanoylamino-3-morpholino-propanol [PDMP] and 1-phenyl-2-palmitoylamino-3-pyrrolidino-propenol [P4])^117,118, of ceramide synthase (fumonisin B1)¹¹⁹, or ceramidase (IS, 2R)-D-erythro-2-(N-myristoylamino)-1-phenyl-propanol [MAPP])¹²⁰ have also found popularity. These are also potentially problematic. The ability of PDMP and its related homologues to raise ceramide has recently been found to reside at a second site of action, the inhibition of 1-O-acylceramide synthase⁴¹. De novo ceramide formation, although operative in some systems⁷¹, is generally not the basis for ceramide accumulation. Thus, even if highly specific, this fungal toxin may not alter endogenous ceramide levels. In general, there is little assurance that, even if ceramide levels can be manipulated by these agents, the ceramide accumulates at a site within the cell at which it would be most biologically active. This consideration would also apply for the use of the bacterial sphingomyelinase or the cell permeant ceramides.

The fourth and final evidence in support of ceramide as an important signaling molecule is that some of the same responses can be replicated in yeast systems. Yeasts have been particularly valuable models for understanding sphingolipid biology. A signaling role for ceramides as activators of protein phosphatase 2A¹²¹ and for their role as mediators of the yeast heat shock response¹²² has been garnered from the use of this model. More recently, an important role for sphingosine-1-phosphates has been identified by use of S. cervesiae, first as the basis for the identification of the sphingosine kinase¹²³ and sphingosine-1-phosphate lyase genes¹²⁴, and second through manipulation of long-chain base phosphate levels within yeast. These latter studies have demonstrated a marked protective effect of these metabolites in models of heat shock and nutrient deprivation¹²⁵.

Several mediators have been implicated in the regulation of signaling sphingomyelinases. Diacylglycerol was perhaps the first lipid to be suggested to regulate TNF-–induced ceramide formation. In this model, the hydrolysis of phosphatidylcholine was thought to be upstream from an acidic sphingomyelinase¹²⁶. Support for this was based largely on the use of D609, believed at the time to be a specific inhibitor of the phosphatidylcholine-specific phospholipase C. This observation has engendered some discussion because in other systems sphingomyelin hydrolysis and diacylglycerol formation occur simultaneously, D609 may not be specific for phospholipase C, and protein kinase C activation can, at times, abrogate biological effects thought to be secondary to ceramide generation, such as apoptosis.

Another lipid mediator of sphingomyelinase may be arachidonic acid, the product of phospholipase A2⁷⁹. In HL60 cells, arachidonate formation precedes sphingomyelin hydrolysis. The effect was reproducible by mellitin, an activator of phospholipase A₂, and is absent in a cell line defective in cytosolic phospholipase A₂. The specificity of this effect for arachidonate remains to be determined, since other polyunsaturated fatty acids may activate neutral sphingomyelinase. In addition, whether this effect is specific for the cytosolic phospholipase A₂ or whether other phospholipases may mediate this effect is unknown.

Glutathione, a tripeptide, serves a major role in protecting cells against oxidative stress. Reduced glutathione is an antioxidant that is depleted in cell death. Both reduced and oxidized glutathione have been identified as inhibitors of the neutral magnesium-dependent sphingomyelinase^127,128. Support for a role for glutathione in the regulation of ceramide formation is based on experiments demonstrating that prevention of sphingomyelin hydrolysis and apoptosis can be prevented by repletion of intracellular glutathione.

The interaction of specific membrane proteins coupling receptors that activate sphingomyelinases with the lipases has been proposed. One such protein is factor associated with neutral sphingomyelinase (FAN)¹²⁹. This is a novel WD (tryptophase-aspartic acid) repeat protein suggested to link the neutral sphingomyelinase with the mitogen-activated protein kinase pathway.

CELLULAR TARGETS FOR SPHINGOLIPIDS

Several effectors have been demonstrated for sphingolipid activities. Perhaps the best characterized effector is the set of orphan receptors termed Edg-1, Edg-3, and H218/Edg-5¹³⁰. These comprise a family of structurally homologous receptors with high affinity and specificity for sphingosine-1-phosphate. Edg-3 and Edg-5 are 44% similar to each other and 50% similar to Edg-1. As seven transmembrane-spanning proteins, these receptors are G-protein coupled. Two related receptors, Edg-2 and Edg-4, appear to couple to lysophosphatidic acid¹³¹. Edg-1 was named as the "endothelial differentiation gene" and was identified as an early response gene in endothelial cell differentiation. These receptors are ubiquitously expressed in multiple tissues.

Several signaling events are observed following sphingosine-1-phosphate binding to cells. These include the activation of mitogen-activated protein kinases¹³², phospholipase C¹³³, phospholipase D¹³⁴, and potassium channels¹³⁵. In addition, there is inhibition of adenylyl cyclase¹³³ and mobilization of intracellular calcium¹³⁶. Of potential importance is the observation that both P-cadherin transcription and serum response activation can be induced with sphingosine-1-phosphate¹³⁷. The latter response is rho dependent and mediates stress-fiber formation. Several other responses to sphingosine-1-phosphate have been observed, including mitogenesis, inhibition of ceramide-induced apoptosis, inhibition of cell motility, neurite retraction, and platelet activation.

Recent work has attempted to determine which cellular responses are coupled to which receptor and subunits. At present, it appears that Edg-1 couples to members of the Gi family. Edg-3 and Edg-5 couple to Gi in addition to Gq and G13¹³⁸. The diversity of receptors and G-protein interactions not only explains the relative sensitivity of some but not all sphingosine-1-phosphate signaling to inhibitors such as pertussis toxin, but may clarify the differing cellular responses to this agonist.

Sphingosine-1-phosphate is formed intracellularly and was originally identified as a putative lipid second messenger. It can mobilize calcium from intracellular stores¹³⁹. Sphingosine phosphate kinase activity is stimulated by factors such as phorbol ester¹⁴⁰, platelet-derived growth factor¹⁴¹, and Fc receptor ligation¹⁴². Inhibition of its activity by N,N-dimethylsphingosine inhibits cell growth and promotes apoptosis¹⁴³. The kinase has been cloned in yeast and mammalian cells. There is a high degree of conservation of its primary sequence between yeast, worms, and mammals. There is also a significant degree of similarity with diacylglycerol . The lyase and phosphatases have also been cloned from yeast. Deletion of these genes by homologous recombination creates an enhanced resistance to heat shock and nutrient deprivation.

Several potential cytosolic effectors of ceramide activity have also been identified. One target is one or more cytosolic phosphatases termed "ceramide-activated protein phosphatase" (CAPP)¹⁴⁴. This enzyme is related to the 2A class of heterotrimeric protein phosphatases. Support for this target is observed in yeast in which ceramide-induced growth arrest can be mitigated in yeast lacking the PP2A subunits¹⁴⁵. Protein phosphatase 1 may also be stimulated by ceramide¹⁴⁶.

An important substrate for ceramide-stimulated phosphatase activity may be the retinoblastoma protein. Ceramide and sphingosine stimulate cellular phosphatase activity against Rb protein and induce cell cycle arrest at the G1/S transition¹⁴⁷.

A second activity that has been identified as a ceramide effector has been termed ceramide-activated protein kinase (CAP kinase)¹⁴⁸. This represents a member of the family of proline-directed protein kinases. It has been reported to be the kinase suppressor of ras (KSR) in Drosophila¹⁴⁹. It is suggested to directly activate Raf and mitogen-activated protein kinase activation in response to TNF-.

A third potential effector for ceramide is protein kinase C ¹⁵⁰. Ceramide has been reported to directly bind to this protein kinase C. A common feature for all of these putative targets is the presence of a cysteine-rich domain that may constitute a portion of the ceramide-binding site. Additional suggested ceramide targets include the CPP32, a caspase 3-like apoptotic protease and the stress activated/c-jun N-terminal protein kinase.

Historically, it has been difficult to reconcile the role of ceramide as an effector with potential sites of formation in the plasma membrane, cytosol, and acidic lysosomal or endosomal compartments. Recently, using a photo-crosslinking to a ceramide analogue, cathepsin D, an acidic aspartate protease was identified as a potential target of ceramide generated by the acidic sphingomyelinase¹⁵¹.

APOPTOSIS AND SPHINGOLIPID SIGNALING

Of the proposed roles for sphingolipids in cell signaling, none has engendered greater interest and generated greater controversy than the concept that ceramide is a signal for programmed cell death. Two primary observations have been critical in the development of this hypothesis. First, agonists that induce apoptosis (for example, TNF-, irradiation, chemotherapeutic drugs, and CD95) activate cellular sphingomyelinases and raise cell ceramide levels^90,92,96,97. Second, the exogenous addition of cell-permeant ceramides or of bacterial sphingomyelinase induces an apoptotic response¹⁵². Under this hypothesis, agonists in the form of cell stressors or of peptide hormones serve as cell death signals by activating one or more sphingomyelinases and increasing ceramide, which then acts as a critical effector for the apoptotic pathway.

Under the current model for apoptosis, death receptors such as the TNF and CD95 receptors activate death effector domain-containing enzymes termed caspases ("cysteine protease that cleaves after aspartic acid"). Activation is mediated by binding to the cytoplasmic portions of these receptors through adaptor proteins such as TRADD, TRF-2, and FADD. Death receptor-independent stimuli such as etoposide and irradiation induce caspase activation at other sites. "Inducer" caspases (caspase 8) appear to act upstream of the mitochondrial release of cytochrome c into the cytosol, which in association with Apaf-1 and caspase-9 subsequently process and activate executioner caspases (for example, caspase-3, -6, and -7).

The challenge for proponents of ceramide as a mediator of programmed cell death has been to define the sphingolipid signaling events in the context of well-established pathways for apoptosis. Unfortunately, no consistent view for the role or the site of action of ceramide in either the cell death effector-dependent or -independent pathways. The published data are inconsistent in this regard. Many of the critical questions surrounding the properties and biological role of sphingomyelin signaling detailed previously in this article impact on the evaluation of the strength of the ceramide hypothesis. These questions include identifying the critical sphingomyelinase involved in ceramide formation (for example, acidic vs. neutral sphingomyelinase), determining the cellular site for ceramide accumulation, defining the critical effectors of ceramide action, and demonstrating that sphingomyelin hydrolysis and ceramide accumulation occur in a kinetically consistent manner.

Thus, both neutral and acidic sphingomyelinases have been implicated in cell death pathways^153,154. Ceramides formed secondary to exogenous sphingomyelinase (presumably in the outer membrane leaflet), within acidic cellular compartments, and via activation of a neutral sphingomyelinase (presumably within the inner membrane leaflet) have been implicated in the cell death response. Ceramide has been shown to activate effector¹⁵⁵ and executioner caspases¹⁵⁶. Effector caspase activation, however, may only result from the exposure of cells to short-chain ceramides. Finally, apoptotic stimuli have been associated both with very rapid and slowly rising ceramide levels. Perhaps when the role of ceramide in the cell death response is fully delineated will more fundamental questions regarding sphingolipid signaling be answered.

SPHINGOLIPIDS AND THE KIDNEY

With the discovery of new insights into sphingolipid biochemistry, cellular function, and pharmacology, investigators have begun to consider what role sphingolipids may play in renal physiology and pathophysiology.

Gb3 levels accumulate in Anderson-Fabry disease, a lysosomal storage disorder affecting the kidney. In this X-linked disorder, affected males accumulate Gb3 and globobiosylceramide in their kidneys, vascular, hearts, and peripheral nerve secondary to a loss of -galactosidase A activity, resulting in renal failure, cardiovascular disease, and neuropathy. At least 30,000 individuals are believed to suffer from this sphingolipidosis worldwide and 6000 within the United States¹⁵⁷. However, fewer than 100 patients are reported in the U.S. Renal Data System to suffer end-stage renal disease as a result of Fabry disease. This disparity suggests that many patients go undiagnosed.

Gaucher disease, caused by a defect in -glucosidase, is currently treated with enzyme replacement therapy using mannose-terminated recombinant glucosidase¹⁵⁸. A similar strategy has been employed for Fabry disease and is currently under clinical investigation with promising preliminary results¹⁵⁹. Recombinant enzyme replacement, however, is an exceptionally expensive pharmacologic modality caused by the small potential market and significant manufacturing costs. -Glucosidase currently represents the most expensive drug on the market, with yearly costs in excess of $200,000.

An alternative approach is the use of "substrate depletion," first proposed by Radin¹⁶⁰. Here, glycosphingolipid synthesis is blocked by inhibition of a precursor such as glucosylceramide. Support for this approach has been reported using a form of "experimental epistasis." By crossing knockout mice with the Sandhoff phenotype (due to GM2 accumulation) with those defective in GalNAc transferase, Liu et al were able to prolong life and postpone the neurological phenotype in the double homozygous-affected offspring¹⁶¹. Support for substrate depletion has also been reported for mouse models of Tay-Sachs¹⁶² and Sandhoff disease¹⁶³ using N-butyldeoxynojirimycin, a relatively inactive inhibitor of the cerebroside synthase. In these studies, mice were treated with between 1.2 and 4.8 g/kg/day orally with the inhibitor. The mice exhibited some prolongation of life and reversal of the accumulation of ganglioside within their brains. Given the high dose of drug required and the secondary toxicities of N-butyldeoxynojirimycin, the development of more active and less toxic inhibitors is required. Nevertheless, N-butyldeoxynojirimycin is currently under evaluation in phase I and phase II trials for the treatment of Gaucher and Fabry diseases. At the time of submission, the results of these trials had yet to be reported.

In vitro support for the treatment of Fabry disease is based on the ability of two different PDMP homologues to deplete Gb3 levels in lymphoblasts from a patient with Fabry disease¹⁶⁴. Concentrations as low as 10 nmol/L of either 4'-hydroxyphenyl- or ethylenedioxyphenyl-2-decanoylamino-3-pyrrolidino-propanol resulted in greater than 70% depletion of Gb3. By contrast even 1000-fold higher concentrations of N-butyldeoxynojirimycin were unable to significantly lower Gb3 content. Preliminary in vivo studies using -glactosidase A knockout mice suggest that renal, hepatic, and cardiac Gb3 levels can be significantly lowered with prolonged treatment with the ethylenedioxy P4 homologue.

The verocytotoxin elaborated by pathogenic strains of Escherichia coli (O157:H7) associated with hemolytic uremic syndrome and hemorrhagic colitis binds to globotriaosylceramide. The epidemiological basis for hemolytic uremic syndrome was established by Karmali et al¹⁶⁵. Gb3 is expressed on platelets; however, verotoxin does not affect platelet aggregation. It is therefore the cellular distribution of Gb3 within kidney and endothelium of this glycosphingolipid receptor that correlates best with the pathophysiology of this disorder¹⁶⁶. In support of this view is the observation that adult glomeruli poorly express Gb3, whereas Gb3 is highly expressed in the glomeruli of pediatric patients.

Hemolytic uremic syndrome is also prevalent in the older population, and yet glomeruli from normal older patients are verotoxin receptor negative. This may be explained by the observation that Gb3 levels can be induced in response to multiple cytokines (particularly TNF, IL-1, and lipopolysaccharide), suggesting a role for glycolipids in the mediation of inflammatory responses¹⁶⁷. Gb3 can also be induced by viral infection such as with Epstein-Barr virus and HIV. It is a Burkitt lymphoma-associated antigen¹⁶⁸. It is thus possible that viral prodromes may render adult patients susceptible to hemolytic uremic syndromes. Such a syndrome has been reported for HIV¹⁶⁹.

If sphingosine-1-phosphate is produced and released from activated platelets, then one might speculate that thrombotic disorders such as hemolytic uremic syndrome may be mediated in part by sphingosine-1-phosphate activity. Sphingosine-1-phosphate stimulates mesangial cell proliferation. Additionally, the Edg-1 and Edg-3 receptors can be detected in developing mouse kidney (abstract; Arend et al, J Am Soc Nephrol (Suppl)10:466A, 1999). Given the ability of sphingosine-1-phosphate to stimulate smooth muscle contraction, one might speculate that this lipid may mediate much of the renal pathology and hypertension observed in this syndrome.

Conversely, ceramide has been implicated as a potential mediator of vascular relaxation. Cell-permeant ceramides have been demonstrated to attenuate agonist-stimulated vasoconstriction¹⁷⁰. One potential mechanism for this effect may be the induction of endothelial nitric oxide synthase (eNOS) by ceramide. Exogenous ceramides have been reported to increase NO synthase activity as well as the translocation of eNOS from the endothelial membrane to intracellular sites. Bradykinin, a known inducer of eNOS, also raises cell ceramide levels¹⁷¹.

Finally, ceramide levels have been demonstrated to rise in ischemic renal injury. These changes reportedly occur due to both sphingomyelinase activation¹⁷² and de novo synthesis by ceramide synthase¹⁷³. Establishing whether the increased ceramide in fact is a critical signal in the ischemic injury or more trivially whether it increases secondary to generalized lipid degradation will require additional studies. The delayed kinetics of ceramide formation in many models of apoptosis and its more likely role in regulating executioner caspases will need to be confirmed in models of renal ischemia. In addition, better characterization of the multiple isoforms of the neutral sphingomyelinase and the development of sphingomyelinase-specific inhibitors will be critical to proving this concept. However, if and when a role for ceramide in mediating renal injury is established, this will provide important and obvious strategic approaches to intervention.

CONCLUDING THOUGHTS

Recent studies in the biochemistry and cellular biology of sphingolipids have provided several potentially exciting opportunities for providing new insights into normal and aberrant renal function. In many cases, there exists substantial evidence for a critical role of the unique lipids, such as receptors for bacterial toxin binding and ligands for the Edg family of receptors. In other cases, debate persists, and further work is required to establish proof of concept. Such proof will undoubtedly progress as additional enzymes in the metabolic pathways of sphingolipids are sequenced and characterized, as extracellular and intracellular targets of sphingolipid messengers are unequivocally identified, and as new inhibitors of these enzymes and glycolipid receptors become available.

I have attempted to provide a balanced and critical appraisal of the current status of sphingolipid biochemistry and function. In many cases, the importance of sphingolipids is well established and unequivocal. In other cases, although many proposed functions for sphingolipids are very exciting in their implications in cell biology and pathophysiology, they remain unproven since they are fundamentally descriptive. The potential role of sphingolipids in protein sorting through lipid rafts, as mediators of injury and repair and as receptors for pathogens, provides fertile ground for work by the renal investigative community. Classic models of renal function and disease not only provide particularly useful opportunities for clarifying the role of sphingolipids in the kidney and vasculature, but also provide the opportunity to establish fundamentally and unequivocally a role for sphingolipids in these vital processes.

References

1. Hokin, LE, Hokin, MR: Evidence for phosphatidic acid as the sodium carrier. Nature 1959 184: 1068–1069,  | PubMed | ISI | ChemPort |
2. Shayman, JA, Radin, NS: The structure and function of renal glycosphingolipids. Am J Physiol 1991 260: F291–F302,  | PubMed | ISI | ChemPort |
3. Shayman, JA: Sphingolipids: Their role in intracellular signaling and renal growth. J Am Soc Nephrol 1996 7: 171–182,  | PubMed | ISI | ChemPort |
4. Karlsson, KA: On the chemistry and occurrence of sphingolipid long-chain bases. Chem Phys Lipids 1970 5: 6–43,  | PubMed | ISI | ChemPort |
5. Radin, NS, Martin, FB, Brown, JR: Galactolipide metabolism. J Biol Chem 1957 224: 499–507,  | PubMed | ISI | ChemPort |
6. Bajjalieh, SM, Martin, TFJ, Floor, E: Synaptic vesicle ceramide kinase: A calcium-stimulated lipid kinase that co-purifies with brain synaptic vesicles. J Biol Chem 1990 264: 14917–14921,
7. Abe, A, Shayman, JA, Radin, NS: A novel enzyme that catalyzes the esterification of N-acetylsphingosine. J Biol Chem 1996 271: 14383–14389,  | Article | PubMed | ISI | ChemPort |
8. Bar, LK, Barenholz, Y, Thompson, TE: Effect of sphingomyelin composition on the phase structure of phosphatidylcholine-sphingomyelin bilayers. Biochemistry 1997 36: 2507–2516,  | Article | PubMed | ISI | ChemPort |
9. White, D: Form and Function of Phospholipids. 1973, Amsterdam, Elsevier Publications, pp 441–481,
10. BARENHOLZ, Y: Physiology of Membrane Fluidity, 1984, edited by SHINITZKY M, Sphingomyelin-lecithin balance in membranes: Composition, structure, and junction relationships, Boca Raton, CRC Press, pp 131–173,
11. Myers, M, Wortman, C, Freire, E: Modulation of neuraminidase activity by the physical state of phospholipid bilayers containing gangliosides Gd1a and Gt1b. Biochemistry 1984 23: 1442–1448,  | Article | PubMed | ISI | ChemPort |
12. Masserini, M, Freire, E: Thermotropic characterization of phosphatidylcholine vesicles containing ganglioside GM1 with homogeneous ceramide chain length. Biochemistry 1986 25: 1043–1049,  | Article | PubMed | ISI | ChemPort |
13. Masserini, M, Palestini, P, Freire, E: Influence of glycolipid oligosaccharide and long-chain base composition on the thermotropic properties of dipalmitoylphosphatidylcholine large unilamellar vesicles containing gangliosides. Biochemistry 1989 28: 5029–5032,  | Article | PubMed | ISI | ChemPort |
14. Fra, A, Masserini, M, Palestini, P, Sonnino, S, Simons, K: A photo-reactive derivative of ganglioside GM1 specifically cross-links VIP21-caveolin on the cell surface. FEBS Lett 1995 375: 11–14,  | Article | PubMed | ISI | ChemPort |
15. Okamoto, T, Schlegel, AM, Scherer, PE, Lisanti, MP: Caveolins, a family of scaffolding proteins for organizing "preassembled signaling complexes" at the plasma membrane. J Biol Chem 1998 273: 5419–5422,  | Article | PubMed | ISI | ChemPort |
16. Brown, DA, London, E: Functions of lipid rafts in biological membranes. Annu Rev Cell Dev Biol 1998 14: 111–136,  | Article | PubMed | ISI | ChemPort |
17. Smart, EJ, Ying, YS, Mineo, C, Anderson, RG: A detergent-free method for purifying caveolae membrane from tissue culture cells. Proc Natl Acad Sci USA 1995 92: 10104–10108,  | PubMed | ChemPort |
18. Iwabuchi, K, Handa, K, Hakomori, S: Separation of "glycosphingolipid signaling domain" from caveolin-containing membrane fraction in mouse melanoma B16 cells and its role in cell adhesion coupled with signaling. J Biol Chem 1998 273: 33766–33773,  | Article | PubMed | ISI | ChemPort |
19. Edidin, M: Lipid microdomains in cell surface membranes. Curr Opin Struct Biol 1997 7: 528–532,  | Article | PubMed | ISI | ChemPort |
20. Jacobson, K, Dietrich, C: Looking at lipid rafts? Trends Cell Biol 1999 9: 87–91,  | Article | PubMed | ISI | ChemPort |
21. Brady, RO, Koval, GJ: The enzyme synthesis of sphingosine. J Biol Chem 1958 233: 26–31,  | PubMed | ISI | ChemPort |
22. Williams, RD, Wang, E, Merrill, AHJ: Enzymology of long-chain base synthesis by liver: Characterization of serine palmitoyltransferase in rat liver microsomes. Arch Biochem Biophys 1984 228: 282–291,  | Article | PubMed | ISI | ChemPort |
23. Stoffel, W, Le Kim, D, Sticht, G: Synthesis and degradation of spingosine bases in Ansenula ciferrii. Hoppe Seyler's Z Physiol Chem 1968 349: 1637–1644,
24. Sribney, M: Enzymatic synthesis of ceramide. Biochim Biophys Acta 1966 125: 242–247,
25. Rother, J, Van Echten, G, Schwarzmann, G, Sandhoff, K: Biosynthesis of sphingolipids: Dihydroceramide and not sphinganine is desaturated by cultured cells. Biochem Biophys Res Commun 1992 89: 14–20,
26. Gatt, S: Enzymatic hydrolysis of sphingolipids. I. Hydrolysis and synthesis of ceramides by an enzyme from rat brain. J Biol Chem 1966 241: 3724–3730,  | PubMed | ISI | ChemPort |
27. Lee, T, Ou, M, Shinozaki, K, Malone, B, Snyder, F: Biosynthesis of N-acetylsphingosine by platelet-activating factor: Sphingosine CoA-independent transacetylase in HL-60 cells. J Biol Chem 1996 271: 209–217,  | PubMed | ISI | ChemPort |
28. Sribney, M, Kennedy, EP: The enzymatic synthesis of sphingomyelin. J Biol Chem 1958 233: 1315–1322,  | PubMed | ISI | ChemPort |
29. Ullman, MD, Radin, NS: The enzymatic formation of sphingomyelin from ceramide and lecithin in mouse liver. J Biol Chem 1974 249: 1506–1513,  | PubMed | ISI | ChemPort |
30. Lipsky, N, Pagano, RE: Sphingolipid metabolism in cultured fibroblasts: Microscopic and biochemical studies employing a fluorescent ceramide analogue. Proc Natl Acad Sci USA 1983 80: 2608–2612,  | PubMed | ChemPort |
31. Van Den Hill, A, Van Heusden, GP, Wirtz, KWA: The synthesis of sphingomyelin in the Morris hepatomas 7777 and 5123D is restricted to the plasma membrane. Biochim Biophys Acta 1985 833: 354–357,  | PubMed | ChemPort |
32. Vos, JP, Dehaas, CGM, Can Golde, LM, Lopes-Gardozo, MJ: Relationships between phosphatidylcholine, phosphatidylethanolamine, and sphingomyelin metabolism in cultured oligodendrocytes. J Neurochem 1997 68: 1252–1260,  | PubMed | ISI | ChemPort |
33. Brady, RO, Bradley, RM, Young, OM, Kaller, HJ: An alternative pathway for the enzymatic synthesis of sphingomyelin. J Biol Chem 1965 240: 3693–3694,  | PubMed | ISI | ChemPort |
34. Ghosh, TK, Bian, J, Gill, DL: Intracellular calcium release mediated by sphingosine derivatives generated in cells. Science 1990 248: 1653–1656,  | PubMed | ISI | ChemPort |
35. Ito, M, Kurita, T, Kita, K: A novel enzyme that cleaves the N-acyl linkage of ceramides in various glycosphingolipids as well as sphingomyelin to produce their lyso forms. J Biol Chem 1995 270: 24370–24374,  | Article | PubMed | ISI | ChemPort |
36. Shayman, JA, Abe, A: Glucosylceramide synthase: Assay and properties. Methods Enzymol 1999 311: 42–49,  | ISI |
37. Coste, H, Martel, MB, Got, R: Topology of glucosylceramide synthesis in Golgi membranes from porcine submaxillary glands. Biochim Biophys Acta 1986 858: 6–12,  | PubMed | ISI | ChemPort |
38. Dressler, KA, Kolesnick, RN: Ceramide 1-phosphate, a novel phospholipid in human leukemia (HL-60) cells: Synthesis via ceramide from sphingomyelin. J Biol Chem 1990 265: 14917–14921,  | PubMed | ISI | ChemPort |
39. Bohnenberger, E, Sandermann, H Jr: Diglyceride kinase from Escherichia coli: Purification in organic solvent and some properties of the enzyme. Eur J Biochem 1979 94: 401–407,  | Article | PubMed | ISI | ChemPort |
40. Hinkovska-Galcheva, VT, Boxer, LA, Mansfield, PJ, Harsh, D: The formation of ceramide-1-phosphate during neutrophil phagocytosis and its role in liposome fusion. J Biol Chem 1998 273: 33203–33209,  | Article | PubMed | ISI | ChemPort |
41. Abe, A, Shayman, JA: Purification and characterization of 1-O-acylceramide synthase, a novel phospholipase A2. J Biol Chem 1998 11: 8467–8474,
42. Stoffel, W, Heimann, G, Hellenbroich, B: Sphingosine kinase in blood platelets. Hoppe Seylers Z Physiol Chem 1973 354: 562–565,  | PubMed | ISI | ChemPort |
43. Stoffel, W: Studies on the biosynthesis and degradation of sphingosine bases. Chem Phys Lipids 1970 5: 139–158,  | PubMed | ISI | ChemPort |
44. Waggoner, DW, Gomez-Munoz, A, Dewald, J, Brindley, DN: Phosphatidate phosphohydrolase catalyzes the hydrolysis of ceramide 1-phosphate, lysophosphatidate, and sphingosine 1-phosphate. J Biol Chem 1997 271: 16506–16509,  | ISI |
45. Igarashi, Y, Hakomori, S: Enzymatic synthesis of N,N-dimethyl-sphingosine: Demonstration of the sphingosine: N-methyltransferase in mouse brain. Biochem Biophys Res Commun 1989 164: 1411–1416,  | Article | PubMed | ISI | ChemPort |
46. Mandon, EC, Ehses, I, Rother, J, Van Echten, G, Sandhoff, K: Subcellular localization and membrane topology of serine palmitoyltransferase, 3-dehydrosphinganine reductase and sphinganine N-acyltransferase in mouse liver. J Biol Chem 1992 267: 11144–11148,  | PubMed | ISI | ChemPort |
47. Morell, P, Radin, NS: Specificity in ceramide biosynthesis from long chain bases and various fatty acyl coenzyme As by brain microsomes. J Biol Chem 1970 245: 342–350,  | PubMed | ISI | ChemPort |
48. Van Helvoort, A, Van't Hof, W, Ritsema, T, Sandra, A, Van Meer, G: Conversion of diacylglycerol to phosphatidylcholine on the basolateral surface of epithelial (Madin-Darby canine kidney) cells: Evidence for the reverse action of a sphingomyelin synthase. J Biol Chem 1994 269: 1763–1769,  | PubMed | ChemPort |
49. Barnholz, Y, Roitman, A, Gatt, S: Enzymatic hydrolysis of sphingolipids. II. Hydrolysis of sphingomyelin by an enzyme from rat brain. J Biol Chem 1966 241: 3731–3737,  | PubMed | ISI | ChemPort |
50. Spence, MW, Byers, DM, Palmer, FB, Cook, HW: A new Zn2+-stimulated sphingomyelinase in fetal bovine serum. J Biol Chem 1989 264: 5358–5363,  | PubMed | ISI | ChemPort |
51. Rao, BG, Spence, MW: Sphingomyelinase activity at pH 7.4 in human brain and a comparison to activity at pH 5.0. J Lipid Res 1976 17: 506–515,  | PubMed | ISI | ChemPort |
52. Chakraborty, G, Ziemba, S, Drivas, A, Ledeen, RW: Myelin contains neutral sphingomyelinase activity that is stimulated by tumor necrosis factor-alpha. J Neurosci Res 1997 50: 466–476,  | Article | PubMed | ISI | ChemPort |
53. Levade, T, Vidal, F, Vermeersch, S, Andrieu, N, Gatt, S, Salvayre, R: Degradation of fluorescent and radiolabelled sphingomyelins in intact cells by a non-lysosomal pathway. Biochim Biophys Acta 1995 12358: 277–287,
54. Nilsson, A: The presence of spingomyelin and ceramide-cleaving enzymes in the small intestinal tract. Biochim Biophys Acta 1969 176: 339–347,  | PubMed | ISI | ChemPort |
55. Duan, RD, Nilsson, A: Purification of a newly identified alkaline sphingomyelinase in human bile and effects of bile salts and phosphatidylcholine on enzyme activity. Hepatology 1997 26: 823–830,  | PubMed | ISI | ChemPort |
56. Schuchman, EH, Suchi, M, Takahashi, T, Sandhoff, K, Desnick, RJ: Human acid sphingomyelinase: Isolation, nucleotide sequence and expression of the full-length and alternatively spliced cDNAs. J Biol Chem 1991 266: 8531–8539,  | PubMed | ISI | ChemPort |
57. Schissel, SL, Keesler, GA, Schuchman, EH, Williams, KJ, Tabas, I: The cellular trafficking and zinc dependence of secretory and lysosomal sphingomyelinase, two products of the acid sphingomyelinase gene. J Biol Chem 1998 273: 18240–18259,
58. Tomiuk, S, Hofmann, K, Nix, M, Zumbansen, M, Stoffel, W: Cloned mammalian neutral sphingomyelinase: Functions in sphingolipid signaling? Proc Natl Acad Sci USA 1998 95: 3638–3643,  | Article&